The second link you provided is the feature request for what you want. I don’t think adding storage-wide flag makes sense; that original feature request (per file type compression) should be implemented instead, if at all.
Ultimately, IIRC if the compressed data is not smaller than original data — it stored uncompressed. Therefore, on one hand, adding support for selective compression by file name or file type shall be trivial. On the other — this will only result in a tiny bit of CPU time savings during backup of new files, at the expense of additional checks for file content/extensions/more things to configure. Is it worth development and configuration time? I feel it isn’t.
In my opinion, the simpler is the software and fewer knobs it has — the better: Fewer opportunities for bugs, and less changes for users screwups, regardless of experience level. For example, Kopia supports per-file-type compression options, among other things, but it’s over-engineered monstrosity I personally don’t trust, and will pick duplicacy over it every time. I like duplicacy specifically because it’s simple and zero configuration.
.